Table 6.
Association between rough and tumble play and health in children and youth.
| Quality Assessment | No. of Participants | Absolute Effect (95% CI, SE) | Quality | ||||||
|---|---|---|---|---|---|---|---|---|---|
| No. of Studies | Design | Risk of Bias | Inconsistency | Indirectness | Imprecision | Other Considerations | |||
| Social competence (age range between 42 months and 11.2 years, data collected over a single session up to 2 years, aspects of social competence were measured using teacher-report questionnaire, peer nominations of popularity and rejection, social cognitive problem solving task, observer rated) | |||||||||
| 5 | Observational studies a | Serious risk of bias b | Serious inconsistency c | No serious indirectness | Serious imprecision d | None | 359 e | r = 0.30; p < 0.05; R = 0.09 | VERY LOW |
| r = 0.28; p < 0.05; R = 0.07 | |||||||||
| r = −0.28, p < 0.05; R = 0.07 | |||||||||
| r = 0.28, p < 0.05, R = 0.07 | |||||||||
| r = −0.32, p < 0.05; R = 0.10 | |||||||||
| r = −0.30, p < 0.05; R = 0.09 f | |||||||||
| r = 0.42, p = 0.37 g | |||||||||
| Year 1: r = 0.22 p < 0.05; r = −0.37, p < 0.01 | |||||||||
| year 2: r = 0.25, p < 0.05 | |||||||||
| Year 1 RTP to 2 social variables: r = 0.28, p < 0.01 h | |||||||||
| r = 0.34, p < 0.05; r = 0.54, p < 0.01 | |||||||||
| B = −0.87, R2 = 0.14, p = 0.03 | |||||||||
| B = 1.39, R2 = 0.32, p = 0.001 | |||||||||
| B = 3.30, R2 = 0.22, p = 0.006 i | |||||||||
| r = 0.30; r = 0.30, p < 0.05 j | |||||||||
| r = 0.56, p < 0.01 k | |||||||||
| Anti-social behavior (age range between 64 months and 13.5 years, data were collected over 8 months up to 22 months, aspects of anti-social behaviour were was measured using direct observation, teacher ratings, and a video behaviour discrimination task) | |||||||||
| 2 | Observational studies l,m | No serious risk of bias n | Serious inconsistency ° | No serious indirectness | Serious imprecision p | None | 176 q | P (RTP leading to aggression) = 0.28%, z = 4.00, p < 0.05 | VERY LOW |
| χ2(40, N = 42) = 8.17, p < 0.004); r = 0.47, p < 0.01 r | |||||||||
| r = 0.29, p < 0.01 | |||||||||
| P (RTP rough leads to aggression) = 2.26%, p < 0.05s | |||||||||
Notes: RTP, rough and tumble play; Social competence: 0 Randomized trials; 5 Observational studies [90,91,92,93,94,95,96]; a Observational studies include 1 longitudinal study [90] and 4 cross sectional studies [91,92,93,94,95,96]. Dewolf [91] was an unpublished graduate thesis; b It is unclear if participants were blinded to the outcomes assessed, and likely that their behaviour was affected by being observed. The research noted that after speaking to the children about their play the children were “distinctly aware of her presence” during later interactions. The outcome assessor (the researcher) was not blinded to the outcomes being assessed [91]; c RTP was not correlated with popularity [96]; For popular children, RTP was not correlated with antisocial behaviour. For rejected children, RTP was not correlated with interpersonal cognitive problem solving [93]; RTP in year 1 was not related to year 2 social problem solving scores for popular or rejected children [95]; RTP was not correlated with social impact, likes most nominations, likes least nominations, antisocial, or film for boys or girls. For girls, RTP also did not correlate with social preference, or interpersonal cognitive problem solving [94]; Boys’ engagement in RTP with other boys was not related to peer-acceptance by girls. Boys’ RTP with mixed-sex peer s was negatively related to peer acceptance by girls and teacher rated social competence [92]; Boys’ RTP chase was negatively correlated with peer nominations of likes least (r = −0.22, p < 0.05), and was not correlated with peer nominations of likes most, social impact, or social preference; RTP rough was negatively correlated with peer nominations of likes most (r = −0.37, p < 0.01) and was not correlated with peer nominations of likes least, social impact, or social preference [90]; d Low median sample size. Moderate number of included studies (N = 5). e Pellegrini [93,94,95] used the same sample. Results are reported separately but participants are counted once. Pellegrini [95] sample had 94 participants at year 1 and 72 participants at year 2; Pellegrini [90] sample consisted of 82 boys; Pellegrini [96] sample consisted of 42 boys; f Boys’ RTP with same sex peers was correlated with acceptance by same sex peers; Boys’ RTP+ pretend play with same sex peers was correlated with acceptance by same sex peers. Boys’ RTP with mixed sex peers was correlated with same sex peer acceptance. Boys’ RTP with same sex peers was related to teacher-rated social competence. Boys’ RTP with same sex peers was related to teacher-rated social competence. Boys’ RTP with mixed sex peers was negatively correlated with other sex peer acceptance and teacher rated social competence [92]; g Positive peer nominations was correlated with proportion of RTP events [91]; h RTP chase correlated with peer nominations of likes least, but not peer nominations of likes most, social impact, or social preference. RTP rough was negatively correlated with peer nominations of likes most, but was not related to peer nominations of likes least, social impact, or social preference [90]; i RTP flexibility was correlated with interpersonal cognitive problem solving (positive and negative solutions respectively). Popularity was not correlated with any aspect of RTP; RTP relative frequency negatively predicted popularity; RTP flexibility accounted for unique variance in the model to predict negative, and positive solutions to an interpersonal cognitive problem, respectively [96]; j For boys, RTP correlated with social preference and interpersonal cognitive problem solving, respectively, but not social impact, likes most or likes least peer ratings; For girls, RTP did not correlate with social preference, social impact, likes most, likes least, interpersonal cognitive problem solving [94]; k For popular children, RTP correlated with interpersonal cognitive problem solving [89]; Antisocial behaviour: 0 Randomized trials; 2 Observational studies [90,93,94]; l Includes 1 longitudinal study [90] and 1 cross sectional study [93,94]; m Pellegrini [90] is a longitudinal study, however only data from year 1 are included. Children in year 2 met age-based exclusion criteria; n It was not possible to blind assessors to outcomes, however assessors were blinded to children’s sociometric and dominance status [90,93,94]; The probability of RTP leading to observer rated aggression for popular children was not significant; For popular children, RTP was not correlated with anti-social behavior [93]; RTP frequency was not correlated with aggression frequency for boys or girls. RTP was not likely to lead to aggression for children in this study. For boys and girls RTP did not correlate with ability to discriminate between RTP and aggression on a film or with anti-social behaviour [94]; RTP (chasing) was not correlated with observed or teacher rated aggression. RTP (rough housing) was not correlated with teacher rated aggression [90]; p The magnitude of the median sample size was low; The magnitude of the number of included studies was small (N = 2); q The total sample includes 1 study of 82 Caucasian boys only [90]. Pellegrini [93,94] participants were from the same study. Results are reported separately but participants are only counted once. r The probability of RTP leading to observer rated aggression within the 3 min observation period was significant for rejected children. RTP was significantly more likely to lead to observer rated aggression for rejected children than with popular children. RTP positively correlated with anti-social behaviour for rejected children [93]. s RTP (rough housing) was correlated with observed aggression. The probability that RTP (rough housing) would lead to aggression within the 3 min observation period was 2.26% [90].